Related Experiment Videos
Estimating the population median by nomination sampling
Journal of the American Statistical Association
|December 1, 1980
Summary
Nomination sampling, which ranks the largest values from random samples, provides a favorable estimate of the population median. This method is practical for historical data with extreme values or when participants influence case selection.
Area of Science:
- Statistics
- Survey Methodology
Background:
- Traditional methods for estimating population medians rely on simple random sampling.
- Extreme value data or participant-driven sampling introduces challenges for standard estimation techniques.
Purpose of the Study:
- To introduce and evaluate nomination sampling as an alternative method for estimating population medians.
- To explore the utility of nomination sampling in scenarios with historical extreme value data or selective respondent participation.
Main Methods:
- Nomination sampling involves selecting the largest values (nominees) from multiple independent random samples.
- These nominees are rank-ordered, and the population median is estimated via interpolation between specific order statistics.
Main Results:
- The nomination sampling estimate demonstrates favorable comparison to the sample median derived from simple random sampling.
- This technique offers a viable approach for estimating population medians when only extreme values are available in historical datasets.
Conclusions:
- Nomination sampling is an effective and practical method for estimating population medians, particularly in specialized data collection contexts.
- The approach addresses limitations of simple random sampling in scenarios involving extreme values or participant-influenced sampling.